The genus Luciobarbus Heckel, 1843 comprises more than 30 species of medium to large size widely distributed along rivers draining into the Persian Gulf and the Mediterranean, Caspian, and Black seas (Doadrio, 1994; Kottelat & Freyhof, 2007). With approximately 20 species, North Africa and the Iberian Peninsula show the highest diversity in Luciobarbus populations, with the genus being one of the main component of the freshwater fish fauna of these regions (Doadrio, 1994; Machordom et al., 1995). Phylogenetic studies using morphological characters, have provided evidence for Luciobarbus monophyly (Doadrio, 1990), but monophyly has not been unambiguously retrieved by molecular markers, due to the position of the genus Capoeta within Luciobarbus (Levin et al., 2012; Yang et al., 2015). The genera Capoeta and Luciobarbus comprise limnetic species, most of which migrate upstream to spawning areas, characterized by the presence of gravel and clear waters (Banarescu, 1999; Doadrio et al., 2011). Although they can survive in a variety of habitats, outside of the reproductive period they are most abundant in downstream areas. Remarkably, in the freshwaters of Morocco rheophilic Luciobarbus species can also be found, which have habitat requirements and morphological traits that differ from other Luciobarbus species (Doadrio, 1990).

The rheophilic species of Luciobarbus are sympatric with limnetic Luciobarbus and Carasobarbus species in rivers of the Atlas Mountains in Morocco (Doadrio, 1994), but their habitat is restricted to riffle areas, while other Atlas Mountain Luciobarbus are more widespread, with preference for slower waters (Doadrio 1990). The rheophilic Luciobarbus in Morocco have been assigned, so far, to two species: Luciobarbus nasus (Günther, 1874) and Luciobarbus magniatlantis (Pellegrin, 1919), which are morphologically characterized, with respect to other Luciobarbus species, by size rarely reaching 20 cm standard length; thick barbels placed at the extreme anterior of the snout; and males with equal-sized nuptial tubercles uniformly distributed over the body (Almaça, 1970; Doadrio, 1990). This difference in morphology placed L. nasus and L. magniatlantis in an unresolved taxonomic position, and they could not be assigned to Luciobarbus when the genus was formally defined on the basis of osteological traits (Doadrio, 1990). Subsequently, molecular studies of mitochondrial genes have unambiguously placed the rheophilic L. nasus in the genus Luciobarbus together with other North African species (Machordom & Doadrio, 2001a, 2001b).

Within rheophilic Luciobarbus from Morocco, the taxonomic position of L. magniatlantis remains unclear. This species has been traditionally considered a junior synonym of L. nasus (Estève, 1947; Lévêque & Daget, 1984). However, several authors reported morphological traits that unequivocally separate the rheophilic Luciobarbus into two morphological types and consider L. magniatlantis as a separate species (Pellegrin, 1939; Almaça, 1970, 1971; Doadrio, 1990).

So far there were no available studies on the population structure of the rheophilic barbs of the four Moroccan basins in which they occur: Moulouya Basin, on the Mediterranean slope, and Oum er Rbia, Tensift, and Kasab basins, on the Atlantic slope (Almaça, 1970; Doadrio, 1994; Borkenhagen & Krupp, 2013). This lack of population studies compounds confusion about the type locality of Luciobarbus nasus and has hindered the taxonomic work. While the type locality, Tensift Basin, of L. magniatlantis is not in question (Pellegrin, 1919), the type locality of L. nasus is unclear, because the two syntype specimens were preserved together with sea fishes from Mogador (Essaouira) on the seacoast of Morocco (Günther, 1874). It is probable that the two syntypes of Barbus nasus came from material collected in the Kasab River near Mogador by K. v. Fritsch and J. Rein in 1872 (Borkenhagen & Krupp, 2013).

We review the taxonomy of rheophilic Luciobarbus from Morocco, investigating morphology and genetic traits of specimens from Moulouya, Tensift, and Oum er Rbia basins, where they have been previously reported (Almaça, 1970). The recognition of two new distinct taxonomic entities, lead us to the description of two new species of Luciobarbus for the freshwaters of Morocco.

Our study of Luciobarbus populations was based on 15 specimens (14 males, one female) from Tensift Basin, 33 specimens (21 males, 12 females) from Oum er Rbia Basin, and 12 specimens (12 males) from Moulouya Basin (Fig. 1; Table 1). Material from Kasab Basin could not be collected despite consecutive efforts over the course of several years. The Kasab River has been intensively modified in recent years by anthropogenic pressure as a consequence of the touristic development of Essaouira (Mogador). Three cyprinid species were collected in the Kasab River during the past two decades: Luciobarbus ksibi (Boulenger, 1905), “Labeobarbus” reinii (Günther, 1874), and Carasobarbus fritschii (Günther, 1874). However, the endemic “L”. reinii and L. nasus from the Kasab River have not been found in the past ten years, and it is likely that the species have disappeared or at least they are extremely rare. Therefore, our study of rheophilic Luciobarbus from the Kasab River was limited to the museum specimens of the two syntypes of Barbus nasus.

The material studied comprised the following specimens and locations: Tensift Basin: 3 specimens from the Reraia River, Moulay Brahim (31.2862, -7.9597), Morocco (voucher numbers: MNCN 280454-280456); 2 specimens from the N'Fiss River, Imaounane (30.9899, -8.2011), Morocco (voucher numbers: MNCN 208172-280173); 7 specimens from the Ourika River, Morocco (voucher numbers: MNHN-IC-1919-0431-0437); Lectotype and Paralectotypes of Barbus magniatlantis Pellegrin, 1919; 3 specimens from N'Fiss River, Kasbah Goundafi (30.9899, -8.2011), Morocco (voucher numbers: MNHN-IC-1919-0438-040). Oum er Rbia Basin: 1 specimen from the Douna River, El Khemis (32.7508, -5.5416) Morocco, (voucher number: MNCN 55094); 2 specimens from the Serrou River, El Herri (32.8276, -5.6152), Morocco (voucher numbers: MNCN 208115-280116); 14 specimens from Arba River, Ksiba (32.5668, -6.0174), Morocco (voucher numbers: MNCN: 54987-55000); 9 specimens from the Oum er Rbia River, El Borj (33.0157, -5.6295), Morocco (voucher numbers: MNCN 208168-208169; MNCN 71972-71978); 7 specimens from the Chbouka River, El Herri (32.8595, -5.6213), Morocco (voucher numbers: MNCN 279704-279707; MNCN 280083; MNCN 279695-279696). Moulouya Basin: 3 specimens from the Melloulou River, Guercif (34.2180, -3.3467 and 34.215, -3.3756), Morocco (voucher numbers: MNCN 286595-286596; MNCN-290831); 9 specimens from the Moulouya River, Ghafoula (34.145, -3.388), Morocco (voucher numbers: MNCN 290832-290840). Kasab Basin: 2 syntype specimens of Barbus nasus Günther, 1874 from the Kasab River preserved at the Natural History Museum (BMNH 1874.1.30.25-26).

All sampling sites (Fig. 1) presented similar riverine morphology, with clear water, rapid current and gravel bottom.

Twenty-four morphometric measurements were taken with digital callipers (0.01 mm), and ten meristic variables were counted with a stereoscopic microscope. The following acronyms were used for morphometric and meristic characters: A, number of anal fin rays; AFH, anal fin height; AFL, anal fin length; APL, anal peduncle length; BL1, first barbel length; BL2, second barbel length; BD, body depth; BLD, body least depth; C, central caudal fin rays; CFL, caudal fin length; CPL, caudal peduncle length; D, dorsal fin rays, DFL dorsal fin length; DFH dorsal fin height; ED, eye diameter; GR, gill rakers (number); HL, head length; IOW, interorbital width; LL lateral line scales; P, pectoral fin rays; PFL, pectoral fin length; PrAD, pre-anal distance; PrDD, pre-dorsal distance; PrOL, pre-orbital length; PrPD, pre-pectoral distance; PrVD, pre-ventral distance; PsOL, postorbital length; PVL, pectoral-ventral length; RSA, scale rows above lateral line; RSB scale rows below lateral line; SL, standard length; V, ventral fin rays; VFL, ventral fin length; VE, Number of vertebrae. The number of vertebrae was obtained by counting on X-ray images of specimens from all sampled populations. Osteological characteristics were investigated through computer tomography (CT) scan and digital dissection using VGStudio MAX v2.2 (Volume Graphics, http://www.volumegraphics.com).

After constructing the measurement matrix, Burnaby's method was used to correct for size effect. The Burnaby method removes the effects of a within-population size-factor from between-group morphometric analyses through an orthogonal projection procedure (Burnaby, 1966; Röhlf & Bookstein, 1987).

All analyses were conducted with the corrected matrix. Morphometric and meristic characters were analysed independently. To assess sexual dimorphism and to identify the variables that contributed most to the variation among populations, two principal component analyses (PCA) were performed using the covariance matrix for morphometric characters. Statistical analyses were carried out using PAST software (Hammer et al., 2001).

For molecular analyses, we obtained specimens of rheophilic Luciobarbus from Oum er Rbia, Tensift, and Moulouya basins along with the limnophilic L. ksibi from Oum er Rbia, Tensift and Kasab basins, (Table 1). Also, the Iberian species of Luciobarbus and L. setivimensis (Valenciennes, 1842) from Algeria were added. Aulopyge hueguelli Heckel, 1843 and Barbus meridionalis Risso, 1827 were selected as outgroups, based on previous phylogenetic analyses (Zardoya & Doadrio, 1999). Total genomic DNA was extracted from fin-clip tissue using the commercial kit Biosprint15 for tissue and blood (Qiagen). For each specimen, the complete region (1140 bp) of the mitochondrial cytochrome b (cytb) was amplified. Primers and protocols used for PCR for cytb followed Machordom & Doadrio (2001b). After checking PCR products on 1% agarose gels, they were purified by ExoSAP-IT™ (USB) and directly sequenced on MACROGEN service using a 3730XL DNA sequencer. Sequences were deposited in the GenBank database under accession numbers KU257523-KU257539.

Phylogenetic analyses were performed using Bayesian inference (BI) implemented in MrBayes v. 3.2 (Ronquist et al., 2012). The Akaike Information Criterion (Akaike, 1973) implemented in jModeltest (Posada, 2008) was used to determine the evolutionary model that best fit the data. In this case, TIM1+G model was selected. Bayesian inference was performed using two independent runs of four Markov Montecarlo coupled chains (MCMC) of 5×10⁶ generations each to estimate the posterior probability distribution. Topologies were sampled every 100 generations, and a majority-rule consensus tree was estimated after discarding the first 10% of generations. The robustness of the clades was assessed using Bayesian posterior probabilities. The average uncorrected p-distances among Luciobarbus populations were calculated for the cytb gene using MEGA package v. 6.0 (Tamura et al., 2013).

COMPARISON OF MORPHOLOGY AMONG POPULATIONS

Due to the sexual dimorphism of Luciobarbus (Doadrio, 1990), and because few female specimens were available, we removed females from the subsequent morphological analyses. Non-parametric Kruskal–Wallis and Mann–Whitney post hoc comparison analyses were used to assess differences in size and body shape among populations (Appendix 1). No differences in SL were found among Luciobarbus populations. However we found significant differences in morphometric measures, with the exception of preventral and preanal distances, anal and dorsal fin lengths, pre and post-orbital distances, and caudal peduncle length (Appendix 1).

The population from Moulouya basin exhibited an elongated body shape with the smallest body depth of any studied population. The dorsal fin was placed slightly more anterior, and the predorsal distance was shorter, than in Tensift and Oum er Rbia populations. All fins in the Moulouya population were larger than the ones present in the remaining populations, with the height of the anal fin significantly bigger than in other populations, in some specimens reaching, or nearly reaching, the anterior rays of the caudal fin when folded. The population of Oum er Rbia Basin possessed the longest anal caudal peduncle. This population had a narrow skull, characterized by the lowest inter-orbital width. The population of Tensift Basin had the smallest head, with the head and prepectoral distances shorter than in Moulouya and Oum er Rbia populations. The deepest caudal peduncle was also found in the Tensift population. The fins in the Tensift population were significantly smaller than in Moulouya and Oum er Rbia populations. Some specimens of the Tensift population, belonging to paralectoypes of Barbus magniatlantis, presented three pairs of barbels, a characteristic that was not observed in any other populations. An overview of the general morphology of the analysed populations is illustrated in Figs. 2 and 3.

The number of scales along the lateral line (median=45), superior transverse line (median=7.5) and inferior transverse line (median=5.5) was significantly lower in the Moulouya population than in the Tensift and Oum er Rbia (Appendix 1). Scale number along the lateral line (median=51) and superior transverse line (median=9.5) were significantly higher in the Oum er Rbia population than in Tensift and Moulouya populations.

The principal component analysis (PCA) divided the specimens into three groups, corresponding to the populations of the three investigated basins (Fig. 4). The eigenvalues of the three first principal components, with the Burnaby-corrected matrix, explained most of the variance (Table 2). The highest eigenvector values (anal fin height and inter-orbital width) were in agreement with results of Kruskal–Wallis and Mann–Whitney analyses (Table 2).

OSTEOLOGICAL FEATURES (APPENDIX 2)

As shown by morphometric analyses, the skull of the Oum er Rbia population was narrower than the one present in Tensift and Moulouya populations. In particular, the ethmoid bone was longer and narrower (Appendix 2-1). In contrast, the kinethmoid bone of the Oum er Rbia population was shorter and more robust than in other populations (Appendix 2-2). The posterior branch of the lachrymal bone was elongated in the Oum er Rbia population and shorter in the Moulouya population (Appendix 2-3). The dentary bone was shorter in the Tensift population than in other populations (Appendix 2-4). The anterior process of the maxilla of the Tensift population was lesser developed than in Oum Er Rbia and Moulouya populations (Appendix 2-5). The number of pharyngeal teeth in Tensift, and Oum er Rbia populations was usually 4.3.2. As in other Luciobarbus species, the juveniles possessed five teeth in the external row. In the Tensift population, the fifth tooth was occasionally retained in the adults. In the Moulouya population, we found pharyngeal teeth in a 4.2.1 or 4.3.1 configuration (Appendix 2-6). The Oum er Rbia population had a thick inner branch of the pharyngeal bone, with reduced pharyngeal lamina. In the Tensift population, the superior branch of the pharyngeal bone was strongly flexed (Appendix 2-7). In Oum er Rbia populations, the last single ray of the dorsal fin was strongly ossified with a maximum width of 12-14.5% (median=13.8) of its height (Fig. 5). In Tensift (6.7-9.5%, median=8.3) and Moulouya (6.2-9.2%, median=7.4) populations, the last single ray of the dorsal fin showed a lower level of ossification (Fig. 5). The Tensift and Moulouya populations showed weaker denticulations in the last single ray of the dorsal fin, but higher denticulate density (Moulouya 3.1-2.4 teeth/mm, median=2.6 teeth/mm; Tensift 2.2-1.8 teeth/mm, median=2.1 teeth/mm) (Fig. 5). The population from Oum er Rbia had scarce and strong denticulations on the last single ray of the dorsal fin at a density of 1.1-1.4 teeth/mm, median=1.2 teeth/mm (Fig. 5).

MOLECULAR DATA

The phylogenetic analysis using BI, with Aulopyge huegelli and Barbus meridionalis as outgroup taxa, revealed two main clades corresponding primarily to the Iberian and African species (Fig. 6). The Iberian group clustered with Luciobarbus setivimensis of Algeria, as was previously reported (Machordom & Doadrio, 2001b). Unexpectedly, the populations of rheophilic Luciobarbus were not monophyletic, and the Moulouya Basin population was clustered with the Iberian group, while the populations from Oum er Rbia and Tensift basins constituted a monophyletic group with Luciobarbus ksibi from the same basins and from its type locality (Kasab River). The phylogenetical position of the Moulouya population within the Iberian group was not resolved.

Genetic distances within Moulouya, Oum er Rbia and Tensift populations of rheophilic barbels ranged from 0.1 to 0.3%. The Moulouya population showed genetic distances from Tensift and Oum er Rbia populations ranging from 8.1 to 8.4%. The genetic distances between Oum er Rbia and Tensift populations was 1.1% (Table 3). Genetic distances between limnophilic L. ksibi and rheophilic barbs of the same basins (Oum er Rbia and Tensift) were from 5.6 to 5.9%. The genetic distances of Moulouya population with respect to other Luciobarbus species were similar or higher than the ones reported to sister species of cyprinid fishes (Doadrio & Carmona, 2003, 2006; Doadrio & Madeira, 2004; Robalo et al., 2005; Doadrio & Elvira, 2007; Domínguez-Domínguez et al., 2007, 2009). In contrast, genetic distances between Tensift and Oum er Rbia population were lower than the ones found in sister species of Luciobarbus (Casal-López et al., 2015).

TAXONOMIC REMARKS

The three studied populations from the Oum er Rbia, Moulouya, and Tensift basins possessed molecular, osteological, and morphometric traits that identified them as different species. The assignment of Oum er Rbia, Moulouya, and Tensift populations to the rheophilic species L. nasus was not possible, as the combination of diagnostic traits of L. nasus was not shared by the studied populations. The combination of the diagnostic traits of all populations and the syntypes of Barbus nasus are showed in Table 4. The population morphologically most similar to L. nasus was that of the Oum er Rbia Basin. Differences between the Oum er Rbia population and L. nasus were found, primarily in the morphology of the last single ray of the dorsal fin, number of scales in the lateral line, morphology of the pharyngeal teeth, kinethmoid bone shape and caudal peduncle depth. The maximum width of the last single dorsal fin ray of the syntype of Barbus nasus was 11.5% of its height, and does not reach the robustness of that of the Oum er Rbia population (12-14.5%, median=13.8) (Fig. 5). The number of scales in the lateral line was 46-44 in Barbus nasus syntypes and 49-53 scales (x=51, median=51) in Oum er Rbia population. Pharyngeal teeth and kinethmoid bone were more robusts in Oum er Rbia than in L. nasus (Appendix 2-2, 2-7). Anal peduncle length was less of two times the BLD while in Oum er Rbia population was always more of two times.

In the absence of molecular data and a greater number of specimens for morphological studies, L. nasus must be considered an endemic species of the Kasab Basin and L. magniatlantis an endemic species of the Tensift Basin.

The populations of Moulouya and Oum er Rbia basins exhibit multiple diagnostic traits differing from those of L. nasus and L. magniatlantis (Table 4).

DESCRIPTION OF LUCIOBARBUS POPULATIONS

The high degree of morphological and genetic differentiation of Luciobarbus populations endemic to the Moulouya and Oum er Rbia basins justifies the consideration of these population as distinct species. No available names for these populations exist, and therefore these are described as new species in the present study.

Luciobarbus guercifensis sp. nov. Doadrio, Perea & Yahyaoui

http://lsid:zoobank.org:act:6DF6332F-F73B-437F-88E1-2F4AC7F15F4F

DIAGNOSIS: Differs from other known Luciobarbus species by the following combination of characters: 45-48 scales along the lateral line (x=45.6, median=45), 6.5-8.5 (x=7.4, median=7.5) above lateral line, and 5.5 below lateral line. The last single ray of the dorsal fin is strongly ossified with the maximum width 6.2-9.2% (median=7.4) of its length. The last single dorsal fin ray is densely denticulated along its length, and, in adults specimens, the number of denticulations exceeds 25 (Fig. 6). The ethmoid bone is wider than its length. Most specimens possess a single pharyngeal tooth in the inner row. The number of vertebrae is 39-41 (x=39.7, n=11). The interorbital distance is 1.7 to 2.2 times (median=2) the eye diameter. The caudal peduncle is longer than the one seen in L. nasus and L. magniatlantis, with depth 2.1-2.6 times the length of the anal peduncle. The posterior segment (manubrium) of the lachrymal bone is short and high. Differences in diagnostic characters among analysed Luciobarbus populations are presented in Table 4.

DESCRIPTION: D III-V 8, A III 5, P I 15-16, V I 6, C 18; LL 45-48 (x=45.6, median=45), RSA 6.5-8.5 (x=7.4, median=7.5), RSB 5.5. Pharyngeal teeth in adults 4.3.1 or 4.2.1. GR 13-15 (x=14.1, median=14), VE 39-41 (x=39.7, median=40). Rarely reaches 200 mm SL. The body is elongated, relative to maximum body depth compared to other Luciobarbus species. Maximum body depth is 16-19% of SL. Head length (x=28.5 mm) is 25-27% of SL and greater than body depth (x=20.1 mm). The skull is wide, with the ethmoid bone width greater than its length. The interorbital distance is 1.7 to 2.2 times eye diameter. Infraorbital bones are narrow and the lachrymal bone is shorter than in Oum er Rbia population as consequence of a shorter manubrium. The height of the manubrium is 20-23% the lachrymal length. Thick lips and barbels display granular appearance. The first pair of barbels is short and located at the anterior extreme of the mouth, reaching the insertion of the second pair of barbels. The second pair of barbels is thick and reaches the posterior edge of the eye. The anterior barbel is 23.2-29.7%, and the second 34.6-38.6%, of HL. The snout is prominent, with preorbital length 10-12.6% of SL. The preorbital length (x=13.6 mm) is larger than the postorbital length (x=12 mm). The dorsal fin is posterior on the body but slightly more anterior than in other rheophilic barbs, with the predorsal distance being 46.2-50% of SL. The profile of the dorsal fin is concave, with the last single ray ossified with more than 25 denticulations (Fig. 5). In males, the anal fin is longer than the one seen in L. nasus and L. magniatlantis with its height 76-92% of APL. The caudal peduncle is less deep than in L. magniatlantis and L. nasus, with a height 9.9-10.9% of SL. The length of the anal caudal peduncle is 2.1 to 2.6 times its height. In males, the pectoral and ventral fins are long with pectoral fins usually reaching the origin of the ventral fins. Males exhibited numerous nuptial tubercles of equal size distributed over the body and fins. The caudal fin length is 16.6-27.3% of SL. Morphometric and meristic measurements for the holotype and paratypes of Luciobarbus guercifensis are presented in Table 5. The colouration of L. guercifensis is silver-yellowish with a paler ventral area (Fig. 2).

DISTRIBUTION: This species is endemic to Moulouya basin, inhabiting Moulouya and Melloul rivers in riffle areas near to the Guercif village (Fig. 1).

ETYMOLOGY: The species name ‘guercifensis’ was selected, as because the species is mainly distributed around the Guercif village in Morocco.

COMMON NAME: We propose the English common name ‘Guercif barbel’ for this new species.

HABITAT AND BIOLOGY: The species inhabits large rivers, mainly in mid-stream, usually in riffle areas. No information exists on the spawning period or reproductive behaviour.

CONSERVATION: Currently, Luciobarbus guercifensis is a rare species that has been found in few places, localized in riffle areas. These riffle areas are becoming scarce in Moulouya River as a consequence of increasing crop irrigation at its headwaters. The water taken for irrigation, as well as the impact of fertilizers and pesticides on water quality in the lower courses has likely been the primary cause of the recent decline of this population. No quantification of the decline in numbers is available. We suggest that this species should be included in the IUCN category of Endangered.

GENETICS: Uncorrected-p distance of mitochondrial cytb gene between Luciobarbus guercifensis and the other analysed species are presented in Table 3. L. guercifensis shows 63 diagnostic positions in the cytb gene with respect to other rheophilic barbels.

Luciobarbus zayanensis sp. nov. Doadrio, Casal-López & Yahyaoui

http://lsid:zoobank.org:act:7A79DAD4-B60A-43F9-8BF6-617B34670E64

DIAGNOSIS: Differs from other known species of Luciobarbus by the following combination of characters: 49-53 scales along the lateral line (x=51, median=51); 8.5-9.5 scales (x=9.3, median=9.5) above lateral line and 5.5-7.5 (x=6.5, median=6.5) below lateral line. The last single ray of the dorsal fin is strongly ossified with a maximum width of 12-14.5% (median=13.9%) of its length. The last single dorsal fin ray is densely denticulate along its length and, in adults, the number of denticulations is smaller than 23 (Fig. 5). The skull is narrow, with the width of the ethmoid bone smaller than its length. The inter-orbital distance is 1.2 to 1.9 times (median=1.6) eye diameter. The pharyngeal teeth formula in adults is 4.3.2. The number of vertebrae is 41-43 (x=42, n=10). Large deep caudal peduncle 2.1-2.6% of APL. Lachrymal bone with narrow posterior segment (manubrium). Differences in diagnostic characters among analysed Luciobarbus populations are presented in Table 4.

DESCRIPTION: D III-V 8, A III 5, P I 15-16, V I 6, C 18; LL 49-53 (x=51, median=51), RSA 8.5-9.5 (x=9.3, median=9.5), RSB 5.5-7.5 (x=6.5, median=6.5). Pharyngeal teeth in adults 4.3.2. GR 13-16 (x=14.5, median=14.5), VE 41-43 (x=42, n=10). Similarly to other rheophilic Luciobarbus, it is a small-sized species, rarely reaching 200 mm SL. Females are larger (x=128.6 mm) than males (x=112.9 mm). The body is elongated with maximum body depth 18-24% of SL in males and 21-24% in females. The head is typically sharp and narrow. The inter-orbital distance is 1.2 to 1.9 times the eye diameter in males and 1.4 to 1.9 times in females. The skull is narrow, with the ethmoid bone longer than wide. The frontal bones are narrower relative to other rheophilic barbs, notably in the anterior part. Infraorbital bones are narrow, and the lachrymal bone is longer than in other rheophilic barbs and possesses a long manubrium. The height of the manubrium is 10-13% the lachrymal bone length. The kinethmoid bone is more robust than the one found in other rheophilic barbs. Thick lips and barbels with granular appearance. The first pair of barbels is short and positioned at the anterior extreme of the mouth, reaching the insertion of the second pair of barbels. The second pair of barbels is thick and reaches the posterior edge of the eye. The anterior barbel length is 20.1-29.9%, and the second 29.7-39.5% of HL. The snout is prominent, with preorbital length 10.9-12.1% of SL in females and 10.6-13.2 in males. The preorbital length (x=14.8 mm in females and x=13.5 mm in males) is greater than the postorbital length (x=13.5 mm in females and x=12.3 mm in males). The pectoral and ventral fins are larger in males than in females. The pectoral fin length is 76-83% (x=80%) of HL in females and 72-90% (x=83%) of HL in males. The ventral fin length is 65.5-74% (x=70%) of HL in females and 65.5-85% (x=78%) of HL in males. The dorsal fin is located posteriorly on the body, with the predorsal distance being 48.9-52.5% of SL in females and 48.5-52.5% in males. The profile of the dorsal fin is concave, with the last single ray of the dorsal fin ossified with few and strong denticulations (Fig. 5). The caudal peduncle is elongated, and the least body depth is 11-11.9% SL in females and 9.5-11.8% SL in males. The length of the anal caudal peduncle is 2.1 to 2.3 times the height of the anal peduncle in females and 2.1-2.6 times in males. Males exhibited numerous equal-sized nuptial tubercles distributed over the body and fins. The caudal fin length is 20-23% of SL in females and 18.4-24.5% of SL in males. Morphometric and meristic measurements for the holotype and paratypes of Luciobarbus zayanensis are represented in Table 6. The coloration of L. zayanensis is brownish with a paler ventral area and some black spots along the body (Fig. 2).

DISTRIBUTION: This species is endemic to Oum er Rbia basin, inhabiting riffle areas with clear waters in the Atlas Mountains (Fig. 1).

ETYMOLOGY: The species name ‘zayanensis’ has been selected because it is mainly distributed around Middle Atlas region, in the area inhabited by the Zayanes shepherds.

COMMON NAME: We suggest the English common name ‘Zayan barbel’ for this new species.

HABITAT AND BIOLOGY: The species inhabits riffle areas in headwaters of large rivers. Spawning takes places from late April to early May.

CONSERVATION: The habitat of L. zayanensis is being transformed by dam and reservoir constructions and by increased pollution linked to agriculture and urban development for tourism. The population of L. zayanensis is in decline, therefore we suggest that this species should be included in the IUCN category of Vulnerable.

GENETICS: Uncorrected-p distances of mitochondrial gene cytb between L. zayanensis and the other analysed species are presented in Table 3. Luciobarbus zayanensis shows 5 diagnostic positions in the cytb gene.